See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/256472256

Judging the Behavior of People we Know: Objective Assessment, Confirmation of

Pre-Existing Views, or Both?

Article  in  Personality and Social Psychology Bulletin · December 2013

DOI: 10.1177/0146167213507287

CITATIONS READS

35 329

3 authors:

Daniel Leising Anne-Marie Brigitte Gallrein

Technische Universität Dresden Technische Universität Dresden
81 PUBLICATIONS   1,717 CITATIONS    5 PUBLICATIONS   76 CITATIONS   

SEE PROFILE SEE PROFILE

Michael Dufner
Universität Witten/Herdecke
62 PUBLICATIONS   2,057 CITATIONS   

SEE PROFILE

Some of the authors of this publication are also working on these related projects:

Moderators of the Effects of Narcissism on Social Outcomes View project

Social processes underlying the joint development of personality and social relationships during university studies View project

All content following this page was uploaded by Michael Dufner on 10 May 2017.

The user has requested enhancement of the downloaded file.

 Personality and Social Psychology
Bulletin http://psp.sagepub.com/

Judging the Behavior of People We Know: Objective Assessment, Confirmation of Preexisting Views, or
Both?
Daniel Leising, Anne-Marie B. Gallrein and Michael Dufner
Pers Soc Psychol Bull 2014 40: 153 originally published online 9 October 2013
DOI: 10.1177/0146167213507287

The online version of this article can be found at:

http://psp.sagepub.com/content/40/2/153

Published by:

http://www.sagepublications.com

On behalf of:

Society for Personality and Social Psychology

Additional services and information for Personality and Social Psychology Bulletin can be found at:

Email Alerts: http://psp.sagepub.com/cgi/alerts

Subscriptions: http://psp.sagepub.com/subscriptions

Reprints: http://www.sagepub.com/journalsReprints.nav

Permissions: http://www.sagepub.com/journalsPermissions.nav

>> Version of Record - Jan 13, 2014

OnlineFirst Version of Record - Oct 9, 2013

What is This?

Downloaded from psp.sagepub.com at Universitaetsbibliothek on January 14, 2014

507287
research-article2013
PSPXXX10.1177/0146167213507287Personality and Social Psychology BulletinLeising et al.

Article

Personality and Social

Judging the Behavior of People We Know:

Psychology Bulletin
2014, Vol. 40(2) 153­–163
© 2013 by the Society for Personality
Objective Assessment, Confirmation of and Social Psychology, Inc
Reprints and permissions:

Preexisting Views, or Both? sagepub.com/journalsPermissions.nav

DOI: 10.1177/0146167213507287
pspb.sagepub.com

Daniel Leising1, Anne-Marie B. Gallrein1,

and Michael Dufner2

Abstract
The present study investigates the relative extent to which judgments of people’s behavior are influenced by “truth” (as
measured by averaged observer-judgments) and by systematic bias (i.e., perceivers’ preexisting views of target persons).
Using data from online questionnaires and laboratory sessions (N = 155), we demonstrate that self- and peer-judgments of
people’s actual behavior in specific situations are somewhat accurate but are also affected by what perceivers thought of the
targets before observing their behavior. The latter effect comprises a general evaluative component (generally positive or
negative views of targets) and a content-specific component (views of targets in terms of specific characteristics, for example,
“restrained”). We also found that friends, but not targets themselves, tend to judge targets’ behaviors more positively than
unacquainted observers do. The relevance of these findings for person perception in everyday life and in research contexts
is discussed.

Keywords
behavior, judgment, accuracy, evaluation, acquaintance

Received October 30, 2012; revision accepted September 8, 2013

A real-life example: A board of assessors meets to see a
candidate present his research, and then, based on their
observations, decide whether the candidate should be
granted a “certificate of eligibility for tenure” (Habilitation,
specific to the academic system in some European coun-
tries). The meeting is public, thus many people who are not
members of the board—including one of the candidate’s
close friends—are also part of the audience. After the talk,
the candidate’s friend approaches some colleagues and
praises the candidate’s performance: “Now, this is how it’s
done!” and so on. It is obvious that he actually believes the
candidate performed really well and must have impressed
everybody in the room quite a bit. Praising the candidate’s
performance for other reasons (e.g., to influence the deci-
sion that is about to be made) would make no sense for him
because the people he talks to have nothing to do with the
formal evaluation of the candidate. Notably, nobody agrees
with him. Most people just nod politely to what he says and
remain silent. Obviously, the other people in the room, who
just saw the candidate for the very first time, agree that the
candidate’s performance was actually quite horrible. How
may such discrepant perceptions of the same person’s
behavior be explained? May the personal relationship
between the candidate and the friend have something to do
with it?
According to West and Kenny’s (2011) Truth and Bias
Model, the variables affecting all sorts of judgments, includ-
ing judgments of persons, can be categorized as either “truth”
or “bias” variables. Truth variables represent what is “real”
or “objectively correct” about a judgment, whereas bias vari-
ables represent any systematic influences apart from that.
The extent to which a truth or bias variable affects judgments
is called the “force” of that variable. In the present study, we
investigate the relative extents to which self- and peer-
judgments of people’s behavior in specific situations are
affected by how the target actually behaved (truth) and by the
preexisting views that the perceivers had of the targets before
observing their behavior (bias).
Truth in Judgments of Behavior
There are at least two distinct ways in which behavior may
be described. On the one hand, one may focus on qualities of
1
Technical University Dresden, Germany
2
Humboldt-University Berlin, Germany
Corresponding Author:
Daniel Leising, Department of Psychology, Technical University Dresden,
01062 Dresden, Germany.
Email: Daniel.Leising@tu-dresden.de

Downloaded from psp.sagepub.com at Universitaetsbibliothek on January 14, 2014

154 Personality and Social Psychology Bulletin 40(2)
behavior that are measurable with great exactness (e.g.,
whether person A decreases or increases her physical dis-
tance from person B). Such qualities of behavior may often
be assessed without relying much on any subjective human
judgments (e.g., by means of a ruler). The downside of such
assessments is that their meaning is often psychologically
ambiguous, and thus needs to be interpreted a posteriori by
the researcher. In everyday life, on the other hand, people
tend to describe their own and others’ behaviors by meaning-
laden terms like, for example, “friendly,” “hasty,” “hostile,”
or “relaxed.” That is, they focus on the psychological rele-
vance of people’s behavior rather than on its physical quali-
ties. This latter kind of behavioral judgment is also the one
that we are dealing with in the present article. However,
determining whether a perceiver’s judgment of a target’s
behavior by means of a meaning-laden term reflects the
“truth” is less straightforward than in the first case.
It may be argued that the best way to assess whether
some behavior is “actually” friendly, hasty, hostile, or
relaxed is the average view that many neutral observers
have of that behavior (cf. Kenny, 2004). Averaging different
perceivers’ judgments of the same behavior reduces the
impact of the individual perceivers’ idiosyncratic term uses
and focuses on the common element instead. The existence of
the latter constitutes the very reason why person-descriptive
language may be effectively used at all. Moreover, averag-
ing judgments of behavior will yield an impression that is
more representative of how a behavior will be perceived by
the average other person. This in turn is important because
how others interpret our behaviors will strongly determine
the consequences of those behaviors (Leising & Müller-
Plath, 2009), so the “social reality” that only exists in other
people’s minds at first (e.g., “the remark that Pete just made
was really offensive”) may eventually have very real effects
on the target person (e.g., Pete may be asked to leave). In
accordance with this view, we define accuracy as the agree-
ment between a given perceiver’s judgment of a given
behavior and the average judgment of the same behavior by
several neutral observers. In terms of West and Kenny’s
(2011) Truth and Bias model, the truth variable in our study
is the average behavior judgment by several neutral observ-
ers, and the extent to which this variable predicts judgments
of the same behavior by any other perceivers (e.g., targets
or informants, see below) is the “force” of the truth
variable.
Using such a definition, research has shown that indi-
viduals are in fact able to judge their own behavior within
specific situations with some degree of accuracy (e.g.,
Leising, 2011; Sadler & Woody, 2003). In these studies,
participants were asked to judge their own behavior during
laboratory interactions, and these self-ratings did predict
judgments of the same interactions by unacquainted observ-
ers. Therefore, we expect that in the present study, self-
judgments of behavior will converge at least somewhat
with judgments by neutral observers. We also address the
Downloaded from psp.sagepub.com at Universitaetsbibliothek on January 14, 2014
previously neglected question of whether the same is true
for behavioral judgments that are provided by close
acquaintances of the targets.
Preexisting Views as Biases in
Judgments of Behavior
The systematic biases that we focus on in the present article
concern perceivers’ preexisting views of the target persons
whose behavior they judge. In terms of the Truth and Bias
model, such preexisting views are bias variables that may
affect behavioral judgments independent of the truth. There
are at least two conceptually distinct ways in which such
views may influence subsequent judgments of behavior, one
of them being globally evaluative and the other being con-
tent-specific. In the first case, a perceiver may simply have
an overall positive (or negative) attitude toward a target, and
thus tend to use all sorts of positive (or negative) terms to
describe that target’s behavior in a given situation, irrespec-
tive of how the target “actually” behaved. In the second case,
a perceiver may have a content-specific view of a target (e.g.,
as being arrogant, witty, or aloof), and thus tend to judge that
target’s behavior in a given situation accordingly, irrespec-
tive of how the target “actually” behaved.
Regarding global evaluation, there is a considerable varia-
tion in how positively or negatively people tend to view them-
selves and others (cf. Bono & Judge, 2003; Furr & Funder,
1998; Kim, Schimmack, & Oishi, 2012; Leising, Erbs, &
Fritz, 2010; Nisbett & Wilson, 1977; Saucier, 1994;
Srivastava, Guglielmo, & Beer, 2010; Wood, Harms, &
Vazire, 2010), and such different attitudes strongly predict
whether perceivers will use positive or negative terms in
characterizing targets (Leising et al., 2010; Leising, Ostrovski,
& Zimmermann, 2013). Sometimes, a perceiver’s character-
ization of a target’s behavior in a given situation may even
depend more on whether the perceiver is fond of the target or
not than on how the target “actually” behaved. We are talking
about partisanship here. Research in social psychology has
demonstrated that perceptions of behavior may be strongly
affected by a perceiver’s loyalty or allegiance regarding the
target person: Studies of presidential debates in the United
States showed that this factor predicted which candidate
viewers would see as the winner (cf. Munro et al., 2002). In
the present study, we expect that a perceiver who already
holds a generalized positive image of a target person will see
the subsequent behaviors of that target person in a more posi-
tive light, whereas a perceiver with a more negative attitude
toward a target will do the opposite. We apply this logic to
self- and peer-judgments of behavior alike.
Apart from global evaluation, a perceiver may think that a
person’s behavior in a given situation was highly (for exam-
ple) sophisticated, friendly, or arrogant because he or she
tended to think of the target as being sophisticated, friendly,
or arrogant even before actually observing the behavior in
question. Note that here it is the specific content of the word,
Leising et al. 155
not its evaluative connotation, that matters. Let us consider,
for example, a perceiver who is supposed to judge a known
target’s “quickness” at a problem-solving task. Regardless of
how positively or negatively the perceiver thinks of the tar-
get in general, if she is convinced that the target is usually
“quick” at accomplishing things, she may tend to judge the
quickness of the target’s problem solving in the situation
accordingly, regardless of how quick the target actually was.
There is some evidence supporting the notion that people
tend to characterize the behaviors of targets in line with pre-
existing images, and that to some extent, this happens inde-
pendent of the targets’ actual behaviors. For example, in the
study by Sadler and Woody (2003), the participants tended to
rate their own behavior during an interpersonal interaction in
accordance with their general self-views, even if their “actual”
behavior during the interaction (as consensually judged by
several observers) was controlled for. In other words, the par-
ticipants overestimated the extent to which their behavior in
the situation concurred with what they generally thought of
themselves. Sadler and Woody (2003) called this phenome-
non a “consistency bias” (cf. Leising, 2011).
In the present article, we will use the term preexisting
views (PEV) to denote the images of the targets that already
existed in the perceivers’ minds before they engage in the
task of judging the targets’ behavior in specific situations. An
important limitation of the aforementioned studies is that
they did not distinguish between preexisting views that are
globally evaluative and preexisting views that are content-
specific. Thus, for example, a given perceiver’s judging a
target’s personality and behavior as “witty” may have been
due to the specific content of the term or simply due to the
fact that the term has a positive evaluative connotation. In the
present study, we disentangle these two possibilities from
one another by separately investigating the influences of pre-
existing views that are globally evaluative (PEV-GE) and
preexisting views that are content-specific (PEV-CS).
Moreover, we also disentangle two components of behav-
ioral judgments that were not disentangled in previous stud-
ies of the “consistency bias”: Preexisting views of targets
may affect subsequent judgments of those targets’ behaviors
because (a) perceivers judge particular targets in certain
ways and/or because (b) perceivers tend to judge all targets
in certain ways. In Kenny’s (1994) terminology, the former
would be denoted as “relationship effects,” whereas the latter
would be denoted as “perceiver effects.” We investigate how
strongly both components contribute to the consistency
between preexisting views and subsequent judgments of
behavior.
Furthermore, in the studies described above the interval
between the global self-assessment of personality and the
momentary self-assessment of situated behavior was very
short (less than an hour). It is thus conceivable that the influ-
ence of the perceivers’ preexisting views was overestimated
because the participants’ general images of themselves were
made salient shortly before they were asked to judge their
Downloaded from psp.sagepub.com at Universitaetsbibliothek on January 14, 2014
own behavior (Leising, 2011), or the personality assessment
took place shortly after the participants had judged their
own behavior (Sadler & Woody, 2003). In the present study,
we use much longer intervals between the two kinds of
assessment in order to rule out this alternative explanation.
Finally, our study extends the previous work by testing
whether the preexisting views affect not only self- but also
peer-judgments of behavior.
Possible Mechanisms
Even though we do not directly address psychological mecha-
nisms in the present article, we think it should still be asked
why one may expect effects of perceivers’ preexisting views
of targets on those perceivers’ subsequent judgments of the
same targets’ behaviors. Would it not be more reasonable for
a perceiver to form as accurate an impression of the target’s
behavior as possible and not rely on any other source of infor-
mation than “what is actually there” (truth)? A large number
of studies have demonstrated that people tend to interpret new
information in accordance with their preexisting beliefs or
expectations (see Nickerson, 1998, for a review). This phe-
nomenon has often been attributed to motivational factors
such as a basic need to maintain cognitive congruence
(Festinger, 1957; Kunda, 1990; Nickerson, 1998). It may be
argued that such congruence is needed because only congru-
ent views permit predictions of future events. With regard to
behavior perception, this would imply that once a perceiver
has formed a certain view of a target person, he or she may be
inclined to actively seek or prioritize novel information that
concurs with this view. In fact, research has shown that once
individuals have developed some belief about another person
(e.g. “the person is friendly”), they selectively seek behav-
ioral evidence that confirms rather than falsifies this belief
(Lord, Lepper, & Preston, 1984; Snyder & Campbell, 1980;
Snyder & Swann, 1978).
Another possible explanation is more informational in
nature. For a perceiver who has access to several sources of
information about a target (e.g., observations of the target’s
behavior in several situations), it may be wise to pool that
information in order to form the most accurate impression of
the target possible. The upside of doing so would be that the
resulting image of the person would likely become more rep-
resentative of how the person is on average, possibly permit-
ting more accurate predictions of the target’s behavior in the
long run. Thus, if the target’s behavior in a new situation
deviates from what the perceiver has previously learned
about the person, the new information may be partially dis-
counted as an “exception,” and the interpretation of the new
behavior may be “colored” by the perceiver’s previous expe-
riences with the target. The downside of doing so, however,
would be that the perceiver’s ability to judge the target’s
behavior in just one situation objectively may at least partly
be compromised. The present study investigates the extent to
which this is the case.
156 Personality and Social Psychology Bulletin 40(2)
Overview of the Present Study
Given the ubiquity and importance of self- and peer-judg-
ments of behavior in everyday life as well as in psychologi-
cal research, it seems worthwhile to examine the factors that
contribute to the formation of such judgments. In the present
study, we use a design that allows us to simultaneously
investigate the influence of truth and bias (i.e., perceivers’
preexisting view of targets) on self- and peer-judgments of
behavior. In order to maximize external validity (cf.
Baumeister, Vohs, & Funder, 2007), we do not use written
vignettes of people’s behavior (which was the predominant
method in previous studies addressing similar issues) but
rather investigate judgments of people’s actual behaviors.
Specifically, we invite target persons and their peers to the
laboratory and then videotape the targets’ behaviors in a vari-
ety of standardized situations. Afterward, the targets and
their peers judge the targets’ behaviors. In order to obtain an
accuracy criterion, the same behaviors are also judged by
unacquainted observers. The general images that the targets
and their peers have of the targets’ personalities are assessed
by means of online questionnaires several days prior to the
laboratory assessments. The design enables us to compare
the relative influence of the truth (= averaged observer-judg-
ments), as compared to PEV-GE and PEV-CS, on the partici-
pants’ judgments of their own and their peers’ behaviors.
Method
Sample
The main sample comprised two groups of participants, the
target persons (N = 155) and the informants (N = 155), who
came to the lab together in dyads (see below). Each partici-
pant (i.e., target or informant) received 15 Euros as a reward
for participating. One hundred and fifty target persons com-
pleted the online questionnaire. Of these, 86 were female, 63
were male, and 1 did not report sex. The mean age of the
targets was 23.2 years (SD = 4.06). Of the informants, 153
completed the online questionnaire. Of these, 93 were
female, 59 were male, and 1 did not report sex. The mean age
of the informants was 23.8 years (SD = 4.44). The majority
(n = 85) of the informants were friends of the targets’.
Another large proportion of the informants were the targets’
romantic partners (n = 53). Other kinds of relationships
between informants and partners (e.g., colleagues, siblings)
were less frequent (n < 10). The informants reported know-
ing the targets quite well (M = 4.17, SD = 0.79, on a scale
ranging from 1 [“not at all”] to 5 [“very well”]) and liking
them very much (M = 4.70, SD = 0.51, on a scale ranging
from 1 [“not at all”] to 5 [“very much”]). The values for the
targets’ knowing and liking of the informants were virtually
identical (Knowing: M = 4.21, SD = 0.73; Liking: M = 4.67,
SD = 0.55), which is unsurprising given that the roles of tar-
get and informant were assigned at random (see below).
Downloaded from psp.sagepub.com at Universitaetsbibliothek on January 14, 2014
Procedure
Recruitment.  We advertised the study on a campus website
and in local community centers. The announcement of the
study contained an e-mail address to which participants were
asked to write in order to register for the study. When pro-
spective participants contacted the research team, they
received an e-mail containing some broad information
regarding the aims and procedures of the study. The e-mail
also asked the participants to recruit a second person who
knew them reasonably well and would also be willing to take
part in the study. When a prospective participant contacted
the research team again and named a second person who had
agreed to participate, we randomly determined which of the
two persons would serve as the target and which would serve
as the informant. In the following, we will call these target-
informant pairs dyads.
Personality questionnaires.  Both the target and the informant
of a dyad received an e-mail containing a link to an online
questionnaire. They were asked to log in and describe the
target person’s personality by means of an adjective list (see
below). We emphasized that the two persons should com-
plete the questionnaire independently and that they should
not communicate with each other about their ratings. The
online questionnaire also contained a number of questions
regarding the two participants’ personal backgrounds (e.g.,
socioeconomic status) and their relationship with each other
(e.g., knowing and liking).
Laboratory session.  A few days later, the target and the infor-
mant were invited to come to the lab together for a series of
observational assessments. These assessments all took place
on the same day and comprised three stages. In the first stage,
the target person engaged in four tasks, which were presented
in random order. Task 1 consisted of answering a number of
questions pertaining to general knowledge (e.g., “How high
is Mount Everest?” “How many people live in the Tokyo
area?” “Who wrote the opera ‘Fidelio’?”). Task 2 consisted
of a role-play in which the target person was to call a “neigh-
bor” (actually played by a confederate) and to demand that
the neighbor turns down the volume on her stereo (cf. Borke-
nau, Mauer, Riemann, Spinath, & Angleitner, 2004). In Task
3, the target person was asked to spontaneously tell a brief
story incorporating the terms “Corkscrew,” “Holiday,”
“Catastrophe,” and “Glove Box.” In Task 4, the target person
was asked to (a) sing a song of his or her own choice, (b) tell
a joke of his or her own choice, and (c) pantomime the term
party (which in German does not also mean “political party”
but only “festivity”). The four tasks were selected to enable
an assessment of diverse qualities of behavior. Completing a
task usually took a target less than 90 s. The targets’ behavior
while engaging in the tasks was videotaped. While the target
of a given dyad completed the four tasks, the respective
informant was seated in another room and asked to wait.
Leising et al. 157
In the second stage of the laboratory session, the target and
the informant each rated four video-clips, by means of the
same adjective list that was also used for the personality
assessments (see below). These video-ratings took place in
separate rooms. One member of the dyad (the target or the
informant) rated the video-clips that had just been produced—
the ones showing the current target engage in the four tasks.
The tasks were presented in the same order in which the target
had completed them. Meanwhile, the other member of the
dyad watched four video-clips showing the target of the pre-
vious dyad (i.e., an unknown person) engage in the same
tasks. Both the target and the informant of the current dyad
were asked to watch one of the respective video-clips that had
been assigned to them, then rate the respective target’s behav-
ior by means of the adjectives, then watch the next video-clip,
and so on. Whether the informant watched the current target
and the current target watched the previous target, or vice
versa, was also determined at random.
In the third stage of the laboratory session, the previous
rating assignments were reversed: If the informant had just
rated the current target, and the current target had just rated
the previous target, then the informant was now supposed to
rate the previous target, whereas the current target was sup-
posed to rate himself or herself. If the rating assignments in
the second stage had been the other way round, then the
assignments in the third stage were reversed accordingly.
This way, we let both the informants and the targets judge (a)
targets they knew and (b) targets they did not know.
External observers’ ratings.  Each of the 620 video-clips (= 155
targets × 4 tasks) was also independently rated by at least
three observers who were unacquainted with the targets (the
maximum number of observers per video was six). The over-
all pool of observers comprised 18 persons from which sub-
groups were selected randomly to judge the individual
video-clips. The ratings were balanced such that each observer
would get to see all targets but only see each target in one
randomly chosen situation. Hence, the observer-ratings
always reflected judgments by persons who saw the respec-
tive target in just one situation. The observers did not receive
any training (apart from general technical instructions on how
to watch the videos and provide their ratings) in order to pre-
serve the representativeness of their intuitive ways of apply-
ing the adjectives to the target persons’ behaviors.
Measure
The adjective list that we employed for the ratings of the tar-
gets’ personalities and behavior in the lab was devised by
Borkenau and Ostendorf (1998) and has been used in several
person perception studies (e.g., Leising et al., 2010; Leising et
al., 2013). It consists of 30 terms that assess the Big Five per-
sonality factors by six items each. Three terms for each factor
have a positive valence and three have a negative valence. We
let the participants use the same terms for assessing the targets’
Downloaded from psp.sagepub.com at Universitaetsbibliothek on January 14, 2014
more stable behavioral dispositions (in the online personality
questionnaire) and their more transient situation-bound behav-
iors (in the lab). This conforms to the personality states
approach, which posits that both traits and states may be
assessed by means of the same vocabulary (Fleeson, 2001;
Bleidorn, 2009). For example, in the online questionnaire, the
targets and their informants rated how “shy” or “egoistic” they
considered the targets to be in general, whereas in the lab ses-
sion, they rated how “shy” or “egoistic” they considered the
targets’ behavior during the tasks to be. However, because fac-
tor structures of measures may differ considerably depending
on whether a measure is used for assessing states versus traits
(e.g., Leising & Bleidorn, 2011), we only report analyses at
the level of individual items in the following. This is also the
level of abstraction at which everyday communication about
individual differences in personality and behavior takes place.
The intervals between the completion of the online question-
naire and the laboratory assessments had an average length of
M = 4.08 days (SD = 3.24) for the targets and of M = 4.43 days
(SD = 4.37) for the informants.
Each of the 30 adjectives was also rated in terms of how
positive or negative an impression of a target person it evokes
(i.e., social desirability). For this, we used the ratings that
were provided by 24 student raters as part of the study by
Leising et al. (2010). The scale ranged from 1 (very negative)
to 7 (very positive). Interrater reliability for these ratings was
ICC(3, 24) = .98. The desirability ratings were needed to
compute global evaluation indices by correlating a perceiv-
er’s ratings of a target on the 30 items with the profile of item
desirabilities (Edwards, 1953). The resulting correlations
reflect the extent to which the perceiver describes the target
in a more positive or negative manner overall.
Results
Agreement Between the Observers
We used averaged observer-ratings as a measure of the tar-
gets’ “true” behaviors during the laboratory tasks. The aver-
age interrater agreement, ICC(1, 3), between three observers
in judging the targets’ behavior by means of a single item
was .45 (SD = 0.15) for the story task, .47 (SD = 0.19) for the
pantomime task, .44 (SD = 0.18) for the role-play task, and
.49 (SD = 0.11) for the knowledge task. These levels of inter-
rater agreement are comparable to the levels of agreement
that were obtained in previous studies (e.g., Borkenau et al.,
2004) and can be regarded as quite satisfactory. Note that
they reflect lower-bound estimates for interrater agreement
because the performances of most targets in most situations
were judged by more than three observers.
Correlations Among the Predictors
In West and Kenny’s (2011) Truth and Bias Model, the truth
and bias variables may or may not be correlated with one
158 Personality and Social Psychology Bulletin 40(2)

Table 1.  Prediction of the Targets’ and the Informants’ Behavior Ratings From the Truth and the Two Bias Variables.

Standardized Betas
Without
Criterion Predictor Betas Raw perceiver effects
Behavior rating by target Truth .22 (.16, .28) .14 (.10, .18) .17 (.13, .21)
  PEV-CS bias .18 (.14, .22) .17 (.13, .21) .13 (.10, .15)
  PEV-GE bias .28 (.21, .35) .09 (.06, .11) .03 (.00, .05)
Behavior rating by Truth .31 (.23, .38) .18 (.13, .23) .18 (.15, .22)
informant PEV-CS bias .21 (.17, .25) .19 (.15, .23) .08 (.05, .11)
   PEV-GE bias .22 (.13, .30) .08 (.04, .10) .04 (.02, .06)

Note. Coefficients were first computed for each task and item separately and then averaged (99% bootstrap confidence intervals are in parentheses).
Truth = averaged observer-ratings of the targets’ behavior, PEV-CS = Perceivers’ preexisting views of targets (content-specific) as measured by the
perceivers’ response to the same item in the personality questionnaire. PEV-GE = Perceivers’ preexisting views of targets (global evaluation) as measured
by the correlation between the perceiver’s description of the target in the personality questionnaire and the social desirabilities of the 30 items. PEV-GE
was multiplied with minus 1 when predicting behavior ratings on items with a negative valence (average desirability rating < 4).

another. In our sample, all correlations between the three pre-
dictors (averaged across items) were statistically significant
as indicated by 99% bootstrap confidence intervals that did
not include zero. The targets’ self-ratings on an individual
questionnaire item (PEV-CS) predicted the observers’ ratings
of the targets’ behavior on the same item (truth) at r = .14 on
average. The respective correlation for the informants’ ratings
of the targets’ personalities was r = .11. Thus, the targets’ and
the informants’ item-specific views of the targets’ personali-
ties had some validity, in that they could predict independent
ratings of the targets’ behaviors in the lab later on.
Before computing the average correlation of PEV-GE
with PEV-CS and with the observer-ratings of the targets’
behavior, PEV-GE had to be inverted (i.e., multiplied with
−1) in all cases where it was correlated with an item that had
a negative valence (i.e., an average desirability rating < 4).
Otherwise, the correlations of PEV-GE with positive and
negative items might have canceled out when averaging
across items. After implementing this rule, the average cor-
relation between PEV-GE and PEV-CS was .28 for the tar-
gets’ self-ratings and .34 for the informant-ratings, suggesting
that targets who received more positive overall personality
ratings were also judged more positively on individual items.
The averaged observer-ratings of the targets’ behavior (truth)
correlated at .06 (on average) with PEV-GE in the targets’
self-ratings of their personalities and also at .06 with PEV-GE
in the respective informant-ratings. Thus, targets who had
received more positive overall personality ratings were also
judged in a slightly more positive manner by the average
observer, on individual items.
Relative Influence of Truth and Bias on
Judgments of the Behavior of Known Targets
Using multiple regressions, we predicted the participants’
(i.e., the targets’ and the informants’) judgments of the tar-
gets’ behavior in the lab from (a) averaged observer-judg-
ments of the same behavior (truth), (b) the same perceivers’
preexisting views of the targets in terms of particular items
(PEV-CS), and (c) the same perceivers’ preexisting global
evaluations of the targets (PEV-GE). For these analyses, we
used the targets’ and the informants’ ratings of persons they
knew (i.e., the current targets) because without prior acquain-
tance between target and perceiver, behavior judgments
could not be affected by the latter two factors. Note that by
using multiple regressions, we determined the independent
contributions of each predictor, controlling for the other two
predictors.
Table 1 displays the average Betas that emerged for the
three predictors, along with 99% confidence intervals that
were determined by bootstrapping with the 30 items as cases.
To arrive at these correlations, we first ran multiple regres-
sions for each item and task separately (= 120 regressions),
then averaged the resulting coefficients across tasks for each
item (distinguishing between tasks did not lead to any differ-
ent conclusions), and then averaged across the 30 items. In
computing average Betas for PEV-GE, we again reversed the
direction of the predictor for all items with a negative
valence. Otherwise, positive and negative associations might
have canceled out. In fact, when not inverting PEV-GE in
predictions of negative items, the average standardized Betas
for the individual items correlated at r(28) = .94 (targets) and
r(28) = .87 (informants) with the items’ social desirability
ratings. Thus, the more an item entailed a positive or nega-
tive evaluation, the better the perceivers’ ratings of the tar-
gets’ behavior by means of that item could be predicted from
the same perceivers’ preexisting global evaluations of the
targets (PEV-GE).
In our discussion of the average contributions of the three
predictors, we will concentrate on the standardized Betas
that are displayed in the second and third data column of
Table 1. As can be seen from the entries in the second data
column, all three predictors made significant contributions in
predicting the targets’ and the informants’ judgments of the
targets’ behavior: The targets’ actual behavior (truth)—as
judged by unacquainted observers—and the perceivers’

Downloaded from psp.sagepub.com at Universitaetsbibliothek on January 14, 2014

Leising et al. 159
content-specific preexisting views of the targets (PEV-CS)
both made significant contributions of approximately the
same size. In addition, the perceivers’ globally evaluative
views of the targets (PEV-GE) made another significant, but
somewhat smaller, contribution of their own. The relative
strengths of the three effects were approximately the same
for self- and informant-ratings of behavior.
We also controlled for perceiver effects in order to deter-
mine how much of the influence of the two bias variables
was due to general perceptual tendencies as opposed to the
perceivers’ views of the particular targets at hand. In the
present study, a perceiver’s perceiver-effect was obtained by
averaging his or her judgments of the behavior of the known
target (i.e., himself or herself or the other person in his or her
dyad) and the unknown target (i.e., the target of the previous
dyad). Perceiver-effects were calculated separately for each
of the 30 items and subtracted from the participants’ ratings
of the targets’ behavior (i.e., the DV) before repeating the
item-wise multiple regressions described above. The results
are displayed in the third data column of Table 1: Controlling
for perceiver effects did not change the effect size for the
truth variable (i.e., observer-ratings of the targets’ behav-
ior). It did, however, significantly reduce the influence of the
perceivers’ global evaluations of the targets (PEV-GE) on
both self- and informant-ratings of behavior. Thus, a substan-
tial proportion of the above-reported effect of this predictor
was due to perceiver effects. For self-ratings, the respective
confidence interval almost included zero (the lower confi-
dence limit was minimally larger than zero, which is obscured
in the table because of rounding). Finally, controlling for per-
ceiver effects also lowered the impact of PEV-CS on behav-
ior judgments, with the decrease being considerably stronger
for the informants’ (compared to the targets’) judgments of
the targets’ behavior.
Loss of Validity Due to Target-Perceiver
Relationship?
Given that the perceiver’s preexisting views of the targets
were found to affect their subsequent judgments of the tar-
gets’ behavior, we investigated whether the accuracy of judg-
ments of known targets was lower than the accuracy of
judgments of unknown targets. To this aim, we compared the
accuracy of four kinds of judgments with each other: (a) The
targets’ judgments of their own behavior, (b) the informants’
judgments of the behavior of “their” targets, (c) the targets’
judgments of the behavior of the previous targets, and (d) the
informants’ judgments of the behavior of the previous tar-
gets. Thus, we compared the accuracy of perceivers who
judged targets they already knew (a, b) with the accuracy of
perceivers who judged targets they did not know (c, d). In all
four cases, accuracy was defined as the agreement between
the respective kind of judgment (a, b, c, d) and the average
observer-judgment of the same behavior. If acquaintance
Downloaded from psp.sagepub.com at Universitaetsbibliothek on January 14, 2014
with a target impairs validity, then accuracy should be lower
for “a” and “b” as compared to “c” and “d.”
First, we computed the correlations between each of the
four perspectives and the averaged observer-ratings sepa-
rately for each task and item. Then we averaged across tasks.
Across the 30 items, the mean correlation was r = .18 (99%
bootstrap CI = [.13, .22]) for the current targets’ ratings of
their own behavior (a), r = .22 (CI = [.17, .27]) for the current
informants’ ratings of the current targets’ behavior (b), r =
.26 (CI = [.22, .30]) for the current targets’ ratings of the
previous targets’ behavior (c), and r = .21 (CI = [.16, .26]) for
the current informants’ ratings of the previous targets’ behav-
ior (d). Confidence intervals revealed that only the greatest
difference (.18 vs. .26) was significant, suggesting that the
accuracy of the targets’ judgments was greater when they
judged someone else’s behavior as compared to when they
judged their own behavior.
Global Evaluation Across Different Assessments
Finally, we compared how positively or negatively partici-
pants’ personalities and behaviors were judged by the dif-
ferent types of perceivers. In each case, a perceiver’s overall
evaluation of a target was measured in terms of the within-
person correlation between the perceiver’s description of
the target across the 30 adjectives and the social desirability
ratings for these adjectives. Thus, a positive correlation was
indicative of positive global evaluation. Note that global
evaluation coefficients were computed for the participants’
judgments of the targets’ personalities by means of the
online questionnaire (= PEV-GE) as well as for their judg-
ments of the targets’ behavior in the lab. Each target’s
behavior in the lab was rated by one previously acquainted
target (i.e., the target’s self-rating) and one previously
acquainted informant, as well as by one unacquainted target
and one unacquainted informant. In addition, there were
three to six observer-ratings of each target’s behavior. In
order to ensure comparability between perspectives, we
first computed a global evaluation index for each individual
observer and then averaged across observers. For behavior
ratings, evaluation indices were first computed separately
for each task and then averaged across tasks.
The average global evaluation indices were r = .58 (range
= −.14 to .92) for the targets’ self-descriptions and r = .68
(range = .05 to .93) for the informants’ descriptions of “their”
targets’ personalities in the online questionnaire. The targets’
average evaluation was r = .43 (range = −.62 to .88) for their
own behavior in the lab and r = .45 (range = −.36 to .89) for
other targets’ behavior. The informants’ average evaluation
was r = .54 (range = −.48 to .93) for their “own” targets’
behavior in the lab and r = .40 (range = −.41 to .88) for other
targets’ behavior. Finally, the average evaluation of the tar-
gets’ behavior in the lab by a single neutral observer was r =
.41 (range = −.31 to .77).
160 Personality and Social Psychology Bulletin 40(2)
We first compared the evaluation coefficients for the five
different kinds of judgments of the targets’ behavior in the
lab with one another, using one-way repeated measures
ANOVA. The comparison was statistically significant, F(4,
147) = 12.274, p < .001. Pairwise comparisons with
Bonferroni corrections for multiple testing (α = .01) sug-
gested that the informants’ judgments of their own targets’
behavior were significantly more positive than all other
behavior judgments but that none of the other behavior judg-
ments differed from each other in terms of global evaluation.
In a second analysis, we compared the evaluation coeffi-
cients for the targets’ versus the informants’ judgments
(Factor 1) and for questionnaire versus behavior ratings
(Factor 2) with one another, using two-way repeated mea-
sures ANOVA. Note that in this analysis, we only used judg-
ments within the same dyads, that is, judgments of known
targets. The analysis yielded significant effects for both fac-
tors, target-informant, F(1, 146) = 29.009, p < .001; behav-
ior-personality, F(1, 146) = 94.784, p < .001; but no
significant interaction effect, F(1, 146) = 0.215, p = .643.
Hence, informants viewed targets more positively than tar-
gets viewed themselves, and for both targets and observers,
ratings of overall personality were more positive than ratings
of behavior in the lab.
Discussion
Truth and Bias in Judgments of Behavior
The current study aimed to investigate the relative extents to
which self- and informant-ratings of targets’ behavior in spe-
cific situations reflect “the truth,” as opposed to two different
kinds of bias (i.e., PEV-GE and PEV-CS). Taken together, we
found that the truth and both bias variables had significant and
independent effects on self- and informant-ratings of behavior.
To some extent, such ratings reflect (a) the views that external
observers would have of the same behaviors (truth), (b) the
generally positive or negative views that the perceivers had of
the targets before observing their behaviors (PEV-GE), and (c)
the views that the perceivers had of the targets before observ-
ing their behaviors in terms of specific judgment dimensions
(e.g., “restrained”; PEV-CS). The “forces” of the truth and
PEV-CS variables were about equally strong and stronger than
the force of the PEV-GE variable.
With regard to truth, we found that self- and informant-
ratings of behavior moderately agreed with average judg-
ments of the same behavior by three or more observers. This
largely confirms the findings of previous studies (Leising,
2011; Sadler & Woody, 2003) and implies that momentary
self- and informant-ratings of behavior do in fact concur to
some extent with how the same behavior would be inter-
preted by persons who see the target person for the first time.
If one accepts averaged behavior ratings by several neutral
observers as a valid measure of the behavior’s “true” mean-
ing, then this finding supports the notion that self- and
Downloaded from psp.sagepub.com at Universitaetsbibliothek on January 14, 2014
informant-ratings of a person’s behavior are at least partly
accurate or valid.
Apart from that, however, the present article also demon-
strates that people’s judgments of the behavior of targets they
know are affected by two different kinds of bias: evaluative
and content-specific preexisting views of a person. The for-
mer bias closely resembles the effect of partisanship on judg-
ments of candidates’ performances in presidential debates
(cf. Munro et al., 2002). However, the findings of the present
study pertain more directly to everyday interpersonal rela-
tionships: Preexisting positive or negative attitudes do not
only affect people’s judgments of the debate performances of
presidential candidates they (dis-)like, they also affect peo-
ple’s judgments of their own behavior, and of the behavior of
their close acquaintances, across diverse situations.
Notably, the strength of the PEV-GE bias is strongly mod-
erated by the evaluative connotation of the item that is used
to describe the behavior. The more evaluative the item is, the
larger the (positive or negative) impact of the perceiver’s
global evaluation of the target will be (cf. Leising &
Borkenau, 2011). The present study is already the third to
empirically corroborate this interaction between perceivers’
global evaluations of targets (sometimes assessed as “lik-
ing”) and the evaluativeness of person descriptors (cf.
Leising et al., 2010; Leising et al., 2013). In fact, the correla-
tions that we found between the Betas for PEV-GE and the
item desirability ratings were so high (r > .86) that it may be
reasonable to interpret the latter as directly reflecting the
extent to which items will respond to a perceiver’s more pos-
itive or negative attitude toward a target. This renders it
likely that controlling for item desirability may be an effec-
tive way of addressing socially (un-) desirable responding
(cf. Paulhus, 2002), that is, individual differences between
perceivers in terms of their judging (all or particular) targets
too positively or too negatively. Future research needs to
address this highly intriguing possibility.
The other kind of bias that independently affects behavior
judgments is content-specific: If a perceiver has come to
generally attribute some level of a particular characteristic to
a target person, then his or her judgments of that target’s sub-
sequent behavior in specific situations will partly reflect this
general view, even when controlling for the target’s “actual”
behavior in the situation. This finding clearly stands in line
with the previous research (cf. Leising, 2011; Sadler &
Woody, 2003). Going beyond previous findings, however,
the present study demonstrates that this bias effect still holds
if one controls for general evaluation. It is also the first study
to show that informant-ratings of behavior are just as suscep-
tible to this bias as are self-ratings. Moreover, our study dem-
onstrates that a substantial proportion of the effect of PEV-CS
is target-specific and cannot merely be explained in terms of
perceivers’ general judgment tendencies (cf. Kenny, 1994).
Finally, the present study demonstrates that the effect occurs
even if the interval between the assessment of the perceiver’s
general image of the target and same perceiver’s judgment of
Leising et al. 161
the same target’s behavior in a particular situation is several
days long (about 4 days on average). Therefore, it appears
highly unlikely that the effect is merely rooted in the research-
ers’ making the perceivers’ general images of the targets
salient shortly before the behavioral assessment takes place.
An explanation in terms of motivational or informational
mechanisms (see above) seems more viable. Future research
should compare the validities of such theoretical explana-
tions with one another.
Loss of Validity Due to Acquaintance Between
Target and Perceiver?
We also investigated whether judgments of the behavior of
known targets are less accurate than judgments of the behav-
ior of unknown targets. Using averaged observer-ratings as
an accuracy criterion, we found that the targets were in fact
systematically less accurate in judging their own behavior as
compared to how accurate they were in judging the behavior
of other people they did not know. However, it may not be
concluded that self-judgments of behavior are generally less
accurate than are other-ratings because we did not find sig-
nificant differences in accuracy between self-ratings and
informant-ratings of behavior. Also, the informants were
about equally accurate in judging the behavior of known tar-
gets and unknown targets. Therefore, the present study does
not suggest that prior acquaintance between perceivers and
targets is always detrimental to the perceivers’ accuracy in
judging the targets’ behavior. According to our view, the
finding that the targets were more accurate in judging some-
one else than in judging themselves may be attributed to the
operation of self-protective mechanisms in self-judgments
(e.g., Tice, 1991) as opposed to the targets’ heightened atten-
tion to someone else’s performance in the same situations
that they had just undergone themselves.
It should be noted, however, that the targets and the infor-
mants in the present study judged the targets’ behavior
directly after seeing them on video. This probably made it
relatively easy for them to be accurate because they received
the exact same behavioral information on which the observ-
ers also based their judgments. The fact that, even under such
optimal conditions for accuracy, we still found significant
effects for our bias variables (see above) renders these effects
even more impressive. We suspect that, under conditions
where the interval between the observation and the judgment
of behavior is larger, the impact of the perceivers’ preexisting
views of the targets on their judgments of the targets’ behav-
ior might become even stronger and possibly come to impair
accuracy. This issue needs to be addressed by future research.
Global Evaluation Across Perspectives
We also investigated mean differences in how positively or
negatively the targets’ behaviors and personalities were
described across the various assessments. First, we found
Downloaded from psp.sagepub.com at Universitaetsbibliothek on January 14, 2014
that the informants’ judgments of “their” targets’ behavior in
the lab were more positive than any other judgments of that
behavior. This clearly supports the notion of a “pal-serving
bias” (Leising et al., 2010), that is, partisanship in judgments
of the behavior of others we know and like. It should be
noted that this finding concerns a main effect (people in gen-
eral are partisan to their friends), whereas the above-
described effects of PEV-GE concern individual differences
(people who generally see their targets more positively will
describe those targets’ subsequent behaviors more posi-
tively). It should also be noted that we did not find evidence
for a general self-serving bias in judgments of behavior. One
possible explanation for this discrepancy may be that prais-
ing others is less “taboo” than praising oneself (cf. Gallrein,
Carlson, Holstein, & Leising, 2013).
Second, we found that judgments of personality were
more positive than judgments of behavior in the lab. This
may be explained by our tasks eliciting behaviors that were
actually less “positive” than the behaviors that most people
tend to show in their everyday lives (the latter constituting
the basis for personality judgments). Alternatively, it may be
the case that judgments of behavior from video are more
data-driven than broad retrospective judgments of personal-
ity in a questionnaire. Because the latter kind of judgment is
more abstract, it may permit more target-serving, selective
recall of relevant behavioral episodes.
Broader Implications and Outlook
Which implications do the present results have for behav-
ioral assessment in general? Most importantly, they suggest
that, apart from “truth,” momentary self- and other-judg-
ments of behavior reflect a substantial proportion of system-
atic bias variance. If a target’s behavior is judged by someone
who is previously acquainted with the target, the resulting
judgments will partly reflect the perceiver’s preexisting view
of the target, both in terms of general evaluation and in terms
of the specific judgment dimension at hand. In other words,
if we know how a perceiver generally views a target, we will
be able to partly predict that perceiver’s judgments of that
target’s subsequent behaviors, irrespective of how the target
actually behaves. This might explain some of the stunning
discrepancies that sometimes exist between different per-
ceivers’ views on the same targets’ behaviors (e.g., in our
introductory real-life example).
The bias effects that we found may become a problem
especially when multiple momentary assessments of behav-
ior are averaged (as is often the case in experience sampling
studies). For example, if we assume that people’s self-images
are relatively stable over time, they might be expected to
affect all momentary self-assessments of behavior in a simi-
lar manner. In contrast, the targets’ actual behaviors in the
different situations may contribute less to the average of
momentary self-assessments, to the extent that they vary
over time. Taken to the extreme, it would be conceivable that
162 Personality and Social Psychology Bulletin 40(2)
an average of many momentary self-assessments may only
reflect the target’s overall self-images, but none of the tar-
get’s actual behaviors in specific situations anymore, because
the latter have averaged out. In such a case, the momentary
assessments would become superfluous because a single
overall self-assessment would do the job (of assessing what
people think of themselves) just as well. As the current study
shows, using informant-instead of self-judgments of behav-
ior would not be a solution to this problem because infor-
mant-judgments are just as prone to being biased by
preexisting views as are self-judgments. The question of how
much incremental validity (e.g., in predicting important out-
comes) aggregated momentary self- or informant-judgments
of behavior actually have beyond the respective perceivers’
general images of the targets remains to be answered by
future research.
The present study also shows that people tend to judge the
behavior of their friends, but not their own behavior, in
overly positive ways. It seems appropriate to use the term
overly here because we could show that these judgments
were significantly more positive than judgments of the very
same behaviors, by means of the very same terms, by any-
body else, including neutral observers. The finding is in line
with previous research showing that people tend to view
their close acquaintances more positively than they view
themselves (Barelds & Dijkstra, 2009; Leising et al., 2010;
McNulty, O’Mara, & Karney, 2008; Murray, 1999; Murray,
Holmes, & Griffin, 1996; Swami, Stieger, Haubner, Voracek,
& Furnham, 2009) but—according to our knowledge—the
present study is the first study to actually demonstrate that
informant-ratings of behavior deviate positively from inde-
pendent assessments of “the truth.”
This has several important implications for person assess-
ments in real life: For targets, systematically befriending
people who might be asked to judge their behavior later on
(e.g., as members of the board during job applications or as
reviewers for scientific journals) is likely to pay off in terms
of obtaining better evaluations. Institutions (e.g., employers,
scientific journals) that are interested in assessing targets’
performances objectively, however, should take care that no
particularly positive or negative personal relationships exist
between targets and judges because otherwise the resulting
judgments might reflect the influence of those relationships
quite strongly (e.g., the average difference between the infor-
mants’ global evaluations of the behavior of known vs.
unknown targets was d = .51 in the present study).
An important limitation of our study is that all informants
were relatively fond of the targets. As a consequence, the
influence of PEV-GE was probably underestimated due to
variance restriction (Leising et al., 2010; Peabody &
Goldberg, 1989). It seems likely that the force of this vari-
able would become stronger when including perceivers who
differ more in how favorable or critical an image they have
of their respective targets. With regard to self-ratings, this
may be achieved by including, for example, participants with
Downloaded from psp.sagepub.com at Universitaetsbibliothek on January 14, 2014
more negative self-images (e.g., people with elevated levels
of depression). With such a sample, we would expect global
evaluation to play an even more prominent role in predicting
the targets’ judgments of their own behaviors. When infor-
mant-ratings are used, a more realistic estimate of the global
evaluation effect may be obtained by also recruiting infor-
mants who have no particular loyalties toward, or who even
dislike, their respective target persons. Such recruitment may
be challenging, but would clearly be worthwhile at the same
time, because the resulting samples of targets and perceivers
would reflect the interpersonal world we inhabit (comprising
friends and foes, allies and competitors, love interests, loved
ones and “exes”) more accurately.
Declaration of Conflicting Interests
The author(s) declared no potential conflicts of interest with respect
to the research, authorship, and/or publication of this article.
Funding
The author(s) disclosed receipt of the following financial support
for the research, authorship, and/or publication of this article: This
research was supported by a grant (LE 2151/4-1) to Daniel Leising
provided by the German Research Foundation (DFG).
References
Barelds, D. P. H., & Dijkstra, P. (2009). Positive illusions about
a partner’s physical attractiveness and relationship quality.
Personal Relationships, 16, 263-283. doi:10.1111/j.1475-
6811.2009.01222.x
Baumeister, R. F., Vohs, K. D., & Funder, D. C. (2007). Psychology
as the science of self-reports and finger movements: Whatever
happened to actual behavior? Perspectives on Psychological
Science, 2, 396-403. doi:10.1111/j.1745-6916.2007.00051.x
Bleidorn, W. (2009). Linking personality states, current social roles
and major life goals. European Journal of Personality, 23, 509-
530. doi:10.1002/per.731
Bono, J. E., & Judge, T. A. (2003). Core self-evaluations: A review
of the trait and its role in job satisfaction and job performance.
European Journal of Personality, 17, 5-18. doi:10.1002/
per.481
Borkenau, P., Mauer, N., Riemann, R., Spinath, F. M., & Angleitner,
A. (2004). Thin slices of behavior as cues of personality and
intelligence. Journal of Personality and Social Psychology, 86,
599-614. doi:10.1037/0022-3514.86.4.599
Borkenau, P., & Ostendorf, F. (1998). The Big Five as states: How
useful is the five-factor model to describe intraindividual varia-
tions over time? Journal of Research in Personality, 32, 202-
221. doi:10.1006/jrpe.1997.2206
Edwards, A. L. (1953). The relationship between the judged
desirability of a trait and the probability that the trait will be
endorsed. Journal of Applied Psychology, 37, 69-106.
Festinger, L. (1957). A theory of cognitive dissonance (Vol. 2).
Stanford, CA: Stanford University Press.
Fleeson, W. (2001). Towards a structure- and process-integrated
view of personality: Traits as density distributions of states.
Journal of Personality and Social Psychology, 80, 1011-1027.
doi:10.1037//0022-3514.80.6.1011
 Leising et al. 163
Furr, R. M., & Funder, D. C. (1998). A multimodal analysis of per-
sonal negativity. Journal of Personality and Social Psychology,
74, 1580-1591. doi:10.1037/0022-3514.74.6.1580
Gallrein, A. M. B., Carlson, E. N., Holstein, M., & Leising, D.
(2013). You spy with your little eye—People are “blind” to
some of the ways in which they are consensually seen by
others. Journal of Research in Personality, 47, 464-471.
doi.org/10.1016/j.jrp.2013.04.001
Kenny, D. A. (1994). Interpersonal perception: A social relations
analysis. New York, NY: Guilford.
Kenny, D. A. (2004). PERSON: A general model of interpersonal
perception. Personality and Social Psychology Review, 8, 265-
280. doi:10.1207/s15327957pspr0803_3
Kim, H., Schimmack, U., & Oishi, S. (2012). Cultural differ-
ences in self- and other-evaluations and well-being: A study
of European and Asian Canadians. Journal of Personality and
Social Psychology, 102, 856-873. doi:10.1037/a0026803
Kunda, Z. (1990). The case for motivated reasoning. Psychological
Bulletin, 108, 480-498. doi:10.1037/0033-2909.108.3.480
Leising, D. (2011). The consistency bias in judgments of one’s
own interpersonal behavior. Two possible sources. Journal of
Individual Differences, 32, 137-143. doi:10.1027/1614-0001/
a000046
Leising, D., & Bleidorn, W. (2011). Which are the basic meaning
dimensions of observable interpersonal behavior? Personality
and Individual Differences, 51, 986-990. doi.org/10.1016
/j.paid.2011.08.003
Leising, D., & Borkenau, P. (2011). Person perception, disposi-
tional inferences and social judgment. In L. M. Horowitz & S.
Strack (Eds.), Handbook of interpersonal psychology: Theory,
research, assessment, and therapeutic interventions (pp. 157-
170). Hoboken, NJ: John Wiley & Sons.
Leising, D., Erbs, J., & Fritz, U. (2010). The letter of recommen-
dation effect in informant ratings of personality. Journal of
Personality and Social Psychology, 98, 668-682. doi:10.1037/
a0018771
Leising, D., & Müller-Plath, G. (2009). Person–situation integra-
tion in research on personality problems. Journal of Research
in Personality, 43, 218-227. doi:10.1016/j.jrp.2009.01.017
Leising, D., Ostrovski, O., & Zimmermann, J. (2013). Are we
talking about the same person here? Inter-rater agreement in
judgments of personality varies dramatically with how much
the perceivers like the targets. Social Psychological and
Personality Science, 4, 468-474.
Lord, C. G., Lepper, M. R., & Preston, E. (1984). Considering
the opposite: A corrective strategy for social judgment.
Journal of Personality and Social Psychology, 47, 1231-1243.
doi:10.1037/0022-3514.47.6.1231
McNulty, J. K., O’Mara, E. M., & Karney, B. R. (2008).
Benevolent cognitions as a strategy of relationship mainte-
nance: “Don’t sweat the small stuff.” But it is not all small
stuff. Journal of Personality and Social Psychology, 94, 631-
646. doi:10.1037/0022-3514.94.4.631
Munro, G. D., Ditto, P. H., Lockhart, L. K., Fagerlin, A., Gready,
M., & Peterson, E. (2002). Biased assimilation of sociopolitical
argument: Evaluating the 1996 U.S. Presidential Debate. Basic
and Applied Psychology, 24, 15-26.
Downloaded from psp.sagepub.com at Universitaetsbibliothek on January 14, 2014
View publication stats
Murray, S. L. (1999). The quest for conviction: Motivated cogni-
tion in romantic relationships. Psychological Inquiry, 10, 23-
34. doi:10.1207/s15327965pli1001_3
Murray, S. L., Holmes, J. G., & Griffin, D. W. (1996). The ben-
efits of positive illusions: Idealization and the construction of
satisfaction in close relationships. Journal of Personality and
Social Psychology, 70, 79-98. doi:10.1037/0022-3514.70.1.79
Nickerson, R. S. (1998). Confirmation bias: A ubiquitous phenom-
enon in many guises. Review of General Psychology, 2, 175-
220. doi:10.1037/1089-2680.2.2.175
Nisbett, R. E., & Wilson, T. D. (1977). Telling more than we can
know: Verbal reports on mental processes. Psychological
Review, 84, 231-259. doi:10.1037/0033-295X.84.3.231
Paulhus, D. L. (2002). Socially desirable responding: The evolu-
tion of a construct. In H. Braun, D. N. Jackson, & D. E. Wiley
(Eds.), The role of constructs in psychological and educational
measurement (pp. 67-88). Hillsdale, NJ: Erlbaum.
Peabody, D., & Goldberg, L. R. (1989). Some determinants of fac-
tor structures from personality-trait descriptors. Journal of
Personality and Social Psychology, 57, 552-567. doi:10.1037/
0022-3514.57.3.552
Sadler, P., & Woody, E. (2003). Is who you are who you’re talk-
ing to? Interpersonal style and complementarily in mixed-sex
interactions. Journal of Personality and Social Psychology, 84,
80-96. doi:10.1037/0022-3514.84.1.80
Saucier, G. (1994). Separating description and evaluation in the
structure of personality attributes. Journal of Personality
and Social Psychology, 66, 141-154. doi:10.1037/0022-
3514.66.1.141
Snyder, M., & Campbell, B. (1980). Testing hypotheses about
other people: The role of the hypothesis. Personality and
Social Psychology Bulletin, 6, 421-426. doi:10.1177/
014616728063015
Snyder, M., & Swann, W. B. (1978). Hypothesis-testing processes in
social interaction. Journal of Personality and Social Psychology,
36, 1202-1212. doi:10.1037/0022-3514.36.11.1202
Srivastava, S., Guglielmo, S., & Beer, J. S. (2010). Perceiving oth-
ers’ personalities: Examining the dimensionality, assumed sim-
ilarity to the self, and stability of perceiver effects. Journal of
Personality and Social Psychology, 98, 520-534. doi:10.1037/
a0017057
Swami, V., Stieger, S., Haubner, T., Voracek, M., & Furnham, A.
(2009). Evaluating the physical attractiveness of oneself and
one’s romantic partner: Individual and relationship correlates
of the love-is-blind bias. Journal of Individual Differences, 30,
35-43. doi:10.1027/1614-0001.30.1.35
Tice, D. M. (1991). Esteem protection or enhancement? Self-
handicapping motives and attributions differ by trait self-esteem.
Journal of Personality and Social Psychology, 60, 711-725.
Retrieved from http://dx.doi.org/10.1037//0022-3514.60.5.711
West, T. V., & Kenny, D. A. (2011). The truth and bias model of
judgment. Psychological Review, 118, 357-378. doi:10.1037/
a0022936
Wood, D., Harms, P., & Vazire, S. (2010). Perceiver effects as pro-
jective tests: What your perceptions of others say about you.
Journal of Personality and Social psychology, 99, 174-190.
doi:10.1037/a0019390